Wnt Signalling and Its Impact on Development and Cancer

Home , AXIN1, BCL9, Ultrabithorax, WNT11, WNT2B, WNT3

P e r s p e c t i v e s

Drosophila mutants was awarded the Nobel T i m e l i n e Prize in Physiology or Medicine in 1995 (TIMELINE). Today, the term Wnt is therefore Wnt signalling and its impact on an amalgam of Wg and Int10. There had been some earlier work on Wnt signalling, in ‘precloning’ times, when development and cancer the underlying pathways and mechanisms had not been identified. In the 1930s, viral Alexandra Klaus and Walter Birchmeier insertion was discovered to promote mam‑ mary tumours in laboratory mice (see Ref. 11 Abstract | The Wnt signalling pathway is an ancient system that has been highly for an example). Even earlier, in a famous conserved during evolution. It has a crucial role in the embryonic development experiment conducted in 1924, Mangold and of all animal species, in the regeneration of tissues in adult organisms and in Spemann grafted dorsal lips of the blastopore many other processes. Mutations or deregulated expression of components of from developing newt embryos onto the the Wnt pathway can induce disease, most importantly cancer. The first gene to opposing side of the embryo, inducing a 12 be identified that encodes a Wnt signalling component, Int1 (integration 1), was second body axis — a twin-headed embryo . The cause was the activity of the protein later molecularly characterized from mouse tumour cells 25 years ago. In parallel, the termed Wnt in the transplanted tissue frag‑ homologous gene Wingless in Drosophila melanogaster, which produces ment13. This work was awarded the Nobel developmental defects in embryos, was characterized. Since then, further Prize in 1935. Moreover, in an experiment components of the Wnt pathway have been identified and their epistatic performed in 1902 by Morgan, the simple salt relationships have been defined. This article is a Timeline of crucial discoveries lithium chloride also induced double axes in frog embryos by activating the pathway later about the components and functions of this essential pathway. termed the Wnt signalling pathway14,15.

We know today that Wnt and a handful of developmental genetics, cell biology, cancer The canonical Wnt signalling pathway other signalling systems (Notch, Hedgehog, research, biochemistry and immunology. Following the discovery of Int1, for almost TGFβ (transforming growth factor‑β)–BMP In addition, Wnt research has covered the 10 years most successful research into the (bone morphogenetic protein) and recep‑ spectrum of model organisms, including Wnt pathway was in the developmental tor tyrosine kinases) are major molecular worms, flies, frogs, mice and humans field, before the link to human cancer was mechanisms that control embryonic (Supplementary information S1 (table)), and realized (see next section). Many of the development. These signalling systems oper‑ therefore serves as an example of successful genes in the Wnt pathway, which were ate beyond cell and tissue boundaries, but interdisciplinary research. first discovered to function transiently function as morphogens that are secreted in development, turned out to act as from one cell or tissue type to activate surface Early discoveries oncogenes and tumour suppressors when receptors, signal transduction components In 1982, Roel Nusse and Harold Varmus, deregulated in human cancer. Thus, several and transcription factors in neighbouring then working at the University of of the Drosophila mutants identified by cells or tissues, regulating processes such as California, San Francisco, USA, reported Nüsslein-Volhard and Wieschaus’s genetic cell proliferation, survival or differentiation. that a tumour virus (mouse mammary screen in the late 1970s and early 1980s8,9, During development, the activity of such sig‑ tumour virus, MMTV) induced mammary and in other screens (for example, that of nalling systems is tightly regulated, whereas gland tumours in mice by activating the Perrimon and collaborators16,17), displayed in cancer and other diseases they can escape expression of a hitherto unknown gene that defects in embryonic segmentation — that this control. For example, a signalling they named Int1 (integration 1)1 (TIMELINE). is, segment polarity defects — similar to component that functions transiently during A spontaneous loss-of-function mutation Wingless mutants. These were caused by development might become an oncogene in the mouse, swaying, that lacked the ante‑ mutations in armadillo (β-catenin in ver‑ when it undergoes a gain-of-function muta‑ rior cerebrellum (and was first described tebrates), dishevelled or porcupine genes, as tion. Alternatively, an inhibitor might suffer in 1967 (REF. 2)) was shown to be a mutant shown by the groups of Wieschaus, Nusse, a loss-of-function mutation, lose its ability to allele of Int1 (Refs 3,4). A Drosophila mela- Perrimon and collaborators8,9,17–22. Whereas regulate signalling and lose its functions as a nogaster mutant lacking wings, Wingless wild-type embryos contained segments tumour suppressor. Both types of change in (Wg), was described in 1973 (Ref. 5), and with alternating rows of spikes and naked Wnt signalling have been linked to cancer. this fly gene turned out to be the homo‑ belts, segments in mutants contained only Therefore, a great deal of effort is being logue of mammalian Int1 (Refs 6,7). The spikes. By contrast, mutations such as zeste invested worldwide in developing therapeutic Wg mutation also caused segmentation white 3 (also known as shaggy; encodes agents that function by fine-tuning the Wnt defects in Drosophila embryos, and a glycogen synthase kinase 3β) caused oppo‑ pathway. number of segment polarity gene mutations site phenotypes; that is, completely naked This article describes major milestones in Drosophila had already been produced segments. Epistatic analysis of double that have substantially contributed to and extensively characterized by Nüsslein- mutants in the early 1990s demonstrated our understanding of the Wnt signalling Volhard and Wieschaus8,9. Subsequently, that these segment polarity genes function system. The history of Wnt research reads the developmental phenotypes were traced as components of a newly discovered signal like a survey of disciplines and benchmarks to mutations in components of the Wnt transduction pathway, the canonical Wnt in modern research, drawing together signalling system and the work on these pathway19–25 (TIMELINE).

Timeline | Milestones in the 25 years since the discovery of the first Wnt gene

Wingless and Shaggy Nuclear accumulation of β-catenin was found in were found to trigger colorectal cancers135. opposing changes in stability and Porcupine, a multi-transmembrane protein, was phosphorylation of found to process Wnt ligands54. Armadillo82,83. Int1 was found to be Mutations in the gene β-catenin asymmetry and nuclear location was identical to Wingless in adenomatous polyposis Dishevelled was found to regulate embryonic axis formation in Drosophila coli (APC) were found in identified as an essential Xenopus225,226. melanogaster6,7. patients with FAP and in element in the Wnt sporadic colorectal β-catenin was pathway21,24. Frizzled (FZD), a seven-span-transmembrane Inherited human cancers119,120,223. molecularly cloned224. receptor, was identified as the cell-surface colorectal cancer, familial receptor of Wnt ligands41,42. adenomatous polyposis Injection of Wnt1 mRNA LEF (lymphoid enhancer (FAP), was associated with caused the formation of a factor)–TCF (T-cell factor) APC was found to β-catenin was found to directly interact with LEF– deletions of chromosome secondary body axis in transcription factors were directly interact with TCF transcription factors, to be translocated to 5q21–5q22117,118. Xenopus laevis embryos26. molecularly cloned33–35. β-catenin113,114. the nucleus and to modulate gene expression30-32.

1982 1987 1989 1990 1991 1992 1993 1994 1995 1996 1997

Roel Nusse and New Wnt pathway A dominant mutation The mouse Min APC was found to regulate β-catenin Stabilizing β-catenin mutations Harold Varmus found components were (Min) in mice predisposes mutation was found stability81. were identified in melanoma and that Int1, a mouse found in a screen of to multiple intestinal to be a truncation colon cancer cell lines128–130. protooncogene, was lethal mutations in neoplasia126. mutation in Apc127. The Nobel Prize in Physiology or associated with Drosophila17. Medicine was awarded to Nüsslein- The three-dimensional structure MMTV-induced Accumulation of Shaggy was shown to Volhard, Wieschaus and Lewis for their of β-catenin was determined227. mammary gland Armadillo in Drosophila be a cytoplasmic discoveries of the genetic control of early tumours1. was found to be mediator of Wnt embryonic development in Drosophila. Phosphorylation targets β-catenin regulated by Wingless19. signalling22. to ubiquitylation, involving interaction with the E3 ligase β-TrCP, and to proteasome- dependent degradation89–92.

Axin1 (the classical murine gene Fused) encodes Soluble Frizzleds (SFRP2 and FRZB) were identified as Identification of the homeotic gene an intracellular inhibitor of Wnt signalling36. soluble extracellular antagonists of Wnt ligands71–74. Ubx as the first Wnt target gene107.

Blue boxes indicate discoveries of Wnt pathway components in cancer

Axis duplication in frog embryos (the TOPflash reporter assay, was developed by the basic mechanism and the components of experimental model system of Xenopus lae- Clevers and collaborators and is now used Wnt signalling are conserved between inver‑ vis is used) turned out to be an easy assay to worldwide32. tebrates and vertebrates (Supplementary define and characterize components of the The history of TCF and LEF is interest‑ information S1 (table)). The conservation of Wnt pathway. Injection of vertebrate WNT1 ing because the two factors had already the members of the Wnt pathway and their mRNA into early Xenopus embryos resulted been molecularly cloned by Grosschedl, similar interactions in flies, frogs and mice in duplications of the body axis and twin- Jones, Clevers and collaborators in 1991 also suggested that this pathway is relevant headed embryos, as shown by McMahon before their connection to Wnt signalling to human development and may have a role and Moon26 (TIMELINE). Axis duplication was was recognized. This again illustrates how in human disease. also induced by injections of mRNA for different areas of research have merged33–35. Since the mid 1990s, many more dishevelled, β‑catenin, dominant-negative By contrast, negative regulators of the Wnt components of the Wnt signalling pathway gsk3β or lef1 (lymphoid enhancer factor 1)27–30 pathway, obtained by injections of axin1 and have been discovered, in particular the Wnt (TIMELINE). In addition, biochemical analyses axin2 mRNAs into Xenopus embryos, pro‑ cell surface receptors Frizzled41–43, LRP5 by us and others demonstrated direct inter‑ duced a complete loss of the body axis36–38 (LDL-receptor related protein 5) and LRP6 actions between these components, and in (TIMELINE). Genetic analyses in mice con‑ (known as Arrow in Drosophila)44–46. Once particular demonstrated that β‑catenin — a firmed that Axin1 or β‑catenin control axis again, the combination of genetic analyses in molecule that is already known to interact formation36,39. We should also mention that Drosophila and biochemical work in Xenopus with the cell-adhesion molecule E‑cadherin the Axin1 mutation analysed by Costanini was the main contributor in identifying these — translocates to the nucleus where it and collaborators36 was the classical mouse two types of Wnt receptor (Supplementary binds the transcription factors LEF1 and mutant Fused, described in 1949 (REF. 40), information S1 (table)). The nuclear compo‑ TCF (T-cell factor), thereby converting which was subsequently recognized as a part nents Legless (known in vertebrates as B-cell LEF1 into a transcriptional activator30–32. of the Wnt pathway. The identification and lymphoma 9 (BCL9)) and Pygopus were This finding was crucial in understanding characterization of another negative regula‑ identified by the laboratories of Basler, Bienz, the mechanism by which cytoplasmic Wnt tor of the Wnt pathway, the tumour sup‑ Cadigan and others47–51 (TIMELINE). Legless signals confer changes in gene expression in pressor gene Apc (adenomatosis polyposis and Pygopus are co-activators of β‑catenin– the nucleus. A further assay for determin‑ coli), is discussed in the next section. In TCF signalling in Drosophila; their role in ing Wnt–TCF transcriptional activity, the summary, these findings demonstrated that vertebrates is still incompletely understood.

Identification of the protooncogene MYC as a target of the Wnt pathway106.

The stem cell compartment was depleted LEF1 mutations were associated with in the small intestine of Tcf4 null mice178. sebaceous gland tumours in humans, showing that Wnt–β-catenin signalling 243 Axin 1 and axin 2 were found to AXIN1 and AXIN2 muta- BCL9 and Pygopus is inhibited in these tumours . interact with β-catenin, GSK3β tions were associated were identified as WNT11 was Small-molecular-weight Wntless (Evenness interrupted or Sprinter) and APC and to promote GSK3β- with colon cancer124,125. transcriptional identified as the dependent phosphorylation and activators of Wnt antagonists of the elusive early Wnt of Drosophila was found to promote Wnt 62–64 degradation of β-catenin37,38,122. Arrow, LRP5 and LRP6 signalling in Drosophila oncogenic b-catenin– signal in Xenopus secretion . were identified as co- and vertebrates48-51. TCF complex were embryogenesis240. Dickkopf (DKK) was identified as receptors of Frizzled44–46. identified164. The retromer complex in Caenorhabditis a secreted Wnt inhibitor75. Kremen was identified Sclerostin (SOST) elegans is essential in longe range Wnt The three-dimensional as a DKK receptor that Norrin and R-Spondins was identified as a signalling52,53. Grouchos were identified as structure of the inhibits Wnt signaling by were identified as secreted Wnt corepressors of LEF–TCF β-catenin–TCF complex induction of LRP secreted activators of inhibitor that binds to The transcriptional activator Hyrax transcription factors93–95. was determined170. internalization78. Wnt signalling79,80. LRP5 and LRP6241,242. associates with β-catenin102.

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

A conditional gain-of-function mutation of SFRP silencing and Wnt or Dishevelled is overexpressed during Activating LRP5 Recombinant X-chromosome linked β-catenin in mice causes intestinal polyposis134. FRZ overexpression were carcinogenesis146,236,237. mutations were SFRZ8 protein Wilms tumour gene associated with carcino- associated with was shown to (WTX) is a component Casein kinase 1 (CK1) was found to regulate genesis143,144,147,148,234,235. The extracellular Wnt inhibitory thyroid inhibit tumour of the β-catenin β-catenin function84,228. factor 1 (WIF1) is downregulated in tumours150. growth167. destruction complex138. LRP5 was found to many cancers238. Protein phosphatase 2A (PP2A) interacts with transduce Wnt signals by Dishevelled was found to polymerize at the plasma membrane the -catenin destruction complex and The chaperone Boca in Drosophila β recruitment of axin to the and to recruit axin upon Wnt stimulation99. modulates GSK3β (Glycogen synthase kinase 3β) plasma membrane233. (MESD in mice) was found to 229,230 regulate LDL receptor transport60,61. function . LGR5 was identified as a Wnt target gene and stem cell marker 245 The GPI-anchored heparan sulphate proteoglycan The tyrosine kinase receptor in the intestine . Dally (Knypek in zebrafish) in Drosophila was Derailed in Drosophila (RYK in found to regulate Wnt signalling231,232. mammals) was identified as an Two microRNAs are regulated by the Wnt–β-catenin alternative Wnt receptor239. pathway and control Nodal signalling244. WIF, a secreted protein, interacts with and inhibits Wnt ligands76.

Particular progress has also been made in the deRobertis, Niehrs and Nathans71–76 state, cytoplasmic β‑catenin levels remain identification of components of the secretory (TIMELINE). For instance, DKK1 antagonizes low and LEF and TCF in the nucleus inter‑ branch of the Wnt pathway that function Wnt signalling during head formation act with Grouchos to repress Wnt-specific in or are secreted from neighbouring cells. in mice77. An additional DKK-receptor, target genes93–95 (TIMELINE). Mutations in These include the action of the retromer Kremen, was shown to inhibit the Wnt sig‑ genes that control β‑catenin stability, complex in Wnt secretion52,53, Porcupine nalling pathway by internalization of LRP78. such as those that encode members of the in Wnt processing54, palmitoylation and Moreover, secreted proteins such as Norrin destruction complex (APC or axins), or glycosylation of Wnt ligands54–56, lipoprotein and R‑Spondin were shown to be activa‑ β‑catenin itself, have been associated with particles57,58 and chaperones59–61, the trans‑ tors of the canonical Wnt pathway owing cancer progression (see next section). membrane protein Wntless (also known as to their interaction with Frizzled–LRP In the presence of canonical Wnt ligands Evenness interrupted or Sprinter)62–64 and the receptors79,80. In the mid 1990s, it was (FIG. 1b), LRP5–LRP6 is phosphorylated by secreted heparan sulphate proteoglycans65,66 shown by the groups of Polakis, Nusse, CK1γ and GSK3β96,97 (and possibly other (reviewed in Refs 67–69). Palmitoylation Wieschaus and collaborators81–83 that the protein kinases yet to be identified), and of Wnt ligands is essential for their graded control of β‑catenin stability is crucial in Dishevelled is recruited to the plasma action in development58,70 (reviewed in Wnt signalling (TIMELINE). Therefore, in the membrane, where it interacts with Frizzled Ref. 68). absence of Wnt ligands (FIG. 1a), cytoplasmic receptors and polymerizes with other The known components of the Wnt β‑catenin is recruited into a destruction Dishevelled molecules98,99. Phosphorylation signalling pathway can be assembled as complex, in which it interacts with APC of LRP5 or LRP6 and the formation of the follows (FIG. 1): secreted Wnt proteins (there and the axins, and is N‑terminally phos‑ Dishevelled polymer, as well as internaliza‑ are 19 Wnt genes in the human genome) phorylated by casein kinase 1α (CK1a) tion with caveolin100, serve as mediators bind to Frizzled receptors and LRP5–LRP6 and GSK3β84−88 (TIMELINE). Following for the translocation of axin to the plasma co-receptors in the plasma membrane. phosphorylation, β‑catenin is targeted for membrane and inactivation of the destruc‑ Several inhibitors of this interaction were proteasome-dependent degradation involv‑ tion complex. The inactivation of the identified at the end of the 1990s, including ing interaction with β‑TrCP (β-transducin destruction complex allows the cytoplasmic secreted Frizzled-related proteins (SFRPs), repeat-containing protein), a component stabilization and translocation of β‑catenin Dickkopfs (DKKs) and Wnt inhibitory of the E3 ubiquitin ligase complex89–92 to the nucleus. Many aspects of this current factor 1 (WIF1), by the laboratories of (TIMELINE). Therefore, in the non-activated model for inactivation of the destruction

genes that function in cell differentiation DKK Wnt Wnt (siamois and brachyury), signalling (VEGF SFRP WIF (vascular endothelial growth factor), FGF4 a (fibroblast growth factor 4) and FGF18), proliferation (cyclin D1 and MYC), adhesion (E-cadherin and NRCAM (neuronal cell- emen

Kr FZD adhesion molecule)) and many further genes LRP5 and LRP6 that are components of the Wnt pathway β-TrCP DVL CK1 α itself, demonstrating that Wnt signalling can P P P N N β P autoregulate its activity in a positive and neg‑ C GSK3 APC Ub β-catenin ative manner (for example, Frizzleds, DKKs, Ub in Ax C LRPs, Axin 2, β‑TrCP and LEF–TCF). For a frequently updated overview on the Wnt Groucho pathway and its target genes see Roel Nusse’s Proteasomal LEF– webpage. TCF degradation Although the focus of this Timeline article is Wnt–β-catenin — that is, canonical Wnt signalling — it is important to note that Wnt some Wnt ligands and Frizzled receptors, b and the Dishevelleds, are capable of activat‑ ing a β‑catenin-independent, non-canonical Wnt signalling cascade. The fact that P DVL Dishevelleds are involved in both canonical FZD N C LRP5 and LRP6 and non-canonical signalling was important N C CK1γ for the realization that Frizzled relatives N C 108 GS N C might be Wnt receptors . Examples of A K3 xi β n N C non-canonical Wnt signalling are the planar cell polarity (PCP) pathway and the Ca2+- APC N C CBP dependent Wnt signalling pathway (for Pygo C comprehensive reviews of non-canonical BCL9 N Wnt signalling see Refs 109,110). Many func‑ LEF– TCF MYC CYCD1 AXIN2 tions of non-canonical Wnt signalling have been described. For example, signalling by the PCP pathway in Drosophila and Xenopus Figure 1 | The canonical Wnt–β-catenin pathway. a | In the absence of Wnt ligands, β‑catenin is recruited into the destruction complex with APC (adenomatous polyposis coli)Na andtur thee Re axins.views Following | Cancer embryos results in polarization of cells and N‑terminal phosphorylation of β‑catenin by the kinases CK1α (casein kinase 1α) and GSK3β (glycogen directed cell motility, which is referred to as synthase kinase 3β), and subsequent ubiquitylation by β‑TrCP (β-transducin repeat-containing protein, convergent extension movement. The non- an E3 ubiquitin ligase), β‑catenin is proteasomally degraded. Low cytoplasmic levels of β‑catenin canonical Wnt signalling pathway will not be ensure transcriptional repression of Wnt target genes by recruitment of the corepressor Groucho to discussed further here. Mutations in com‑ LEF (lymphoid enhancer factor)–TCF (T-cell factor) transcription factors. b | In the presence of Wnt ponents of non-canonical Wnt signalling in ligands, LRP5 (LDL-related receptor protein 5) and LRP6 are phosphorylated by CK1γ and GSK3β, and human cancer, that is oncogenes or tumour Dishevelled (DVL) molecules are recruited to the plasma membrane to interact with Frizzled (FZD) suppressor genes, have not been described. receptors and other Dishevelled molecules. Interaction of axin with phosphorylated LRP5, LRP6 and However, it is interesting to note that the the Dishevelled polymer leads to the inactivation of the destruction complex and subsequently to the non-canonical Wnt5a has transforming stabilization of β‑catenin and its translocation to the nucleus. In the nucleus, β‑catenin forms a tran- 111,112 scriptionally active complex with LEF and TCF transcription factors by displacing Grouchos and capacity in cell culture . interacting with co-activators such as BCL9 (B-cell lymphoma 9), Pygopus (Pygo) and CBP (CREB binding protein). CYCD1, cyclin D1; DKK, Dickkopf; SFRP, secreted Frizzled-related protein; Canonical Wnt signalling in cancer P, phosphorylation; Ub, ubiquitylation; WIF, Wnt inhibitory factor 1. Until the end of 1993 there was no overlap between research on Wnt signalling and human cancer. Then, Vogelstein, Kinzler complex by the action of Wnt ligands The identification of the protooncogene and Polakis reported an important bio‑ need to be clarified further. In the nucleus, MYC as a direct transcriptional target of chemical interaction between the tumour β‑catenin forms a transcriptionally active Wnt–β-catenin signalling in 1998 shed light suppressor APC and the Wnt pathway complex with LEF and TCF transcription on the transforming activity of the Wnt component β‑catenin113,114 (FIG. 1; TIMELINE; factors30–32 by displacing Grouchos and pathway in cancer106 (TIMELINE) (see next Supplementary information S1 (table)). interacting with other co-activators such section). However, the first direct target Two types of repeat in APC are essential as BCL9, Pygopus, CBP (CREB-binding gene of β‑catenin–LEF — Ultrabithorax for interaction with β‑catenin — three protein) or Hyrax47–51,101–105 (TIMELINE). CBP in Drosophila — was actually identified by 15-amino-acid and seven 20-amino-acid and Hyrax control gene expression through Bienz107 (TIMELINE). Further Wnt target genes repeats (FIG. 2a) that compete with the cell- chromatin remodelling and by influencing have been discovered in the late 1990s and adhesion molecule E‑cadherin for β‑catenin RNA polymerase II. in the current decade. These include target binding115.

a MCR 30 1 3

15 Somatic 4 2 y (second hit) 5

0 2843

equenc c Fr

10 Germline (first hit) a 30

b 1 2 3 4 5

...EIKEKIGTRSAEDPVSEVPAVSQHPRTKSSRLQG…….SA.….EFSSG………QTPKSPPEHYVQETPLMFSRCTSVSSLDSFESRSIASSVQSEPCSGMVSGIISPSDLPDSPGQTMPPSRSKTPPPPPQTAQTKREVPKNKAPTAEKRES,,,DADTLLHFATESTPDG.. 1263 1587

APC 1 2843 Min a b c Oligomerization SAMP repeat ∆APC1 (Axin1/2 binding) Armadillo repeat EB1/RP1 binding ∆APC2 β-catenin binding (15aa repeat) Microtubule binding b β-catenin binding ..QQS29YLDS33GIHS37GATT41TAPS45LS.. and downregulation DLG/PTP-BL binding β-catenin (20aa repeat) Somatic mutation Germline mutation 1 781 (first hit) ∆Ex3 Figure 2 | Human APC and CTNNB1 mutations are associated with (hot spots 1–5) produce truncated APC proteins (∆APC1)Nature harbouring Reviews | Canc oneer carcinogenesis. a | Histogram indicating the frequency of somatic and or two 20-amino-acid repeats. Germline mutations at b or c that are fol- germline mutations within human APC (adenomatous polyposis coli). lowed by mutations in the MCR result in loss of the wild-type allele and Germline mutations are distributed all over the APC gene, with two pref- produce a truncated APC protein (∆APC2).The Min mutation (nonsense erential mutation sites at codon 1061 (a) and codon 1309 (b), which mutation at codon 850) results in a stable truncated APC protein that produce truncated APC proteins (∆APC1 and ∆APC2) that have been predisposes to multiple intestinal neoplasia in mice. b | Somatic muta- associated with familial adenomatous polyposis (FAP). Somatic mutations tions and deletions (∆) in the 5′ sequence (exon 3) of human CTNNB1 (the in the mutation cluster region (MCR) produce truncated APC proteins gene that encodes β-catenin) that have been associated with human that have been observed in sporadic colorectal cancers. People carrying cancers. DLG, disc large; EB1/RP1, a family of microtubule-associated one mutation in APC typically acquire a second mutation during adoles- proteins; PTP-BL, protein tyrosine phosphatase BL. The data in the histo- cence and are predisposed to benign colorectal cancers. First-hit muta- gram in part a is adapted from Ref. 121. The other data in part a is taken tions at codon 1061 (a) accompanied by somatic mutations in the MCR from RefS 155,219,220.

Adenomatous polyposis, a type of human years later, truncating mutations in APC were mediate the interaction of APC with scaffold colon cancer in which numerous polyps characterized in both patients with FAP and proteins of the β‑catenin destruction form in the epithelium of the large intestine, in frequent sporadic colorectal cancers; the complex, Axin 1 and Axin 2 (REFS 37,122) has been described since the mid eighteenth latter represent approximately 85% of human (TIMELINE). In accordance with this, an Apc century, and its hereditary nature was recog‑ colorectal cancers119,120 (reviewed in Ref. 116). mouse mutant with a truncation mutation nized as early as 1900 (reviewed in Ref. 116). A high frequency of APC mutations were after the region that encodes the first SAMP The first clue to the molecular pathogenesis frameshift, nonsense or splice-site mutations, repeat did not develop tumours123. AXIN1 of colon cancer was the 1987 finding that the which resulted in truncations of about 50% and AXIN2 loss-of-function mutations have rare inherited disease, familial adenomatous of the APC protein (reviewed in Ref. 121) also been detected in rare cases of colorectal polyposis (FAP), was associated with dele‑ (FIG. 2a). However, single APC mutations are cancer124,125. A nonsense mutation of Apc was tions of the specific chromosome region insufficient for the induction of adenomatous also produced in an ENU (the mutagen ethyl 5q21–22 (Refs 117,118) (TIMELINE). Patients polyposis and a second mutation (or ‘second nitrosourea)-treated mouse: the Min (Multiple with FAP develop hundreds to thousands of hit’) is always required: mutation of the intestinal neoplasia) mouse, which develops adenomatous polyps in the colon, and with‑ second APC allele. Many APC mutations adenomatous polyposis126,127 (TIMELINE). The out surgical resection some of these polyps accumulate before the region that encodes Min mutant has since become an important develop to malignant carcinomas. Only two the so-called SAMP repeats — regions that animal model in cancer research.

Box 1 | Wnt signalling in human disease are components of the TGFβ signalling pathway. In a recent mouse tumour model, Mutation and altered expression of components of the canonical Wnt pathway are linked to the double mutation of Apc and Smad4 human diseases other than cancer (reviewed in REFS 141,218): leads to activation of the expression of the • Canonical Wnt signalling is involved in bone malformations. Osteoarthritis — a degenerative chemokine CCL9, which recruits immature disorder of the joints — has been associated with augmented Wnt signalling: polymorphisms in myeloid cells from the tumour stroma and SFRP3 (secreted Frizzled-related protein 3), LRP5 (LDL-receptor related protein 5) and WISP1 promotes tumour progression156. Mutations (Wnt1-induced secreted protein 1), which is a direct Wnt target. Gain-of-function mutations in LRP5 and loss of sclerostin, a secreted Wnt inhibitor, are known to affect the homeostatic of components of the Hedgehog signalling balance of osteoblasts and osteoclasts after birth, which leads to high bone mass. Loss-of- pathway, such as loss of Indian hedgehog, function LRP5 mutations lead to osteoporosis-pseudogliome syndrome (OPPG), which is also contribute to the progression of colorec‑ associated with low bone mass. tal cancer157. It therefore seems that it is the • Loss-of-function mutations of LRP5 are not only linked to low bone mass but also to eye inappropriate activation of developmental defects, such as familial exudative vitreoretinopathy (FEVR). Many cases of FEVR are linked to signalling pathways that are responsible for loss-of-funtion mutations in the Wnt receptor FZ4 (Frizzled 4). and promote tumour progression. • Loss-of-function mutations in WNT3 cause tetra-amelia, the loss of all four limbs. Wnt therapeutics • Acute renal failure and polycystic kidneys are associated with gain-of-function mutations in WNT4 and aberrant expression of the target gene PKD1 (polycystic kidney disease 1). Basic researchers and pharmaceutical and biotechnology companies have been inter‑ • Wnt–β-catenin signalling is involved in cardiogenesis and cardiovascular diseases such as cardiac hypertrophy (increased FZ2), as well as in neurodevelopment and neurodegenerative ested in developing Wnt pathway inhibitors diseases such schizophrenia or Alzheimer disease (WNT1 and LRP6 are thought to be since inappropriate activation of the Wnt involved. pathway was first linked to human cancer in the late 1990s (FIG. 3). It was also realized that a number of existing drugs and recently In a fraction of sporadic colorectal regulation of cell proliferation, such as MYC developed derivatives of non-steroidal human cancers, gain-of-function muta‑ and cyclin D1 (Refs 30,32,106,135–137). anti-inflammatory drugs (NSAIDs; for tions of CTNNB1 (the β-catenin gene) have Loss-of-function mutations also occur example, aspirin, indomethacin, sulindac, been discovered; their effects are to prevent in WTX (X-chromosome-linked Wilms celecoxib, rofecoxib and others), or vitamins phosphorylation, subsequent ubiquitylation, tumour; also known as FAM123B), which (such as vitamin A and D derivatives) and proteasomal degradation of β‑catenin. encodes a recently discovered component seem to target the Wnt pathway, directly or Initially, it was reported in a rather limited of the β‑catenin destruction complex138,139. indirectly, for example by inhibiting the Wnt sampling that about 10% of sporadic color‑ Moreover, uncontrolled Wnt–β-catenin target enzyme cyclooxygenase 2 (COX2) ectal cancers contain activating mutations in signalling that is associated with elevated or activating E‑cadherin158–160 (reviewed in CTNNB1 (Refs 128–130). More recent and β‑catenin levels is also linked to aggressive Refs 161,162). Some of the NSAIDs also far more extensive mutational surveys indi‑ fibromatosis and pulmonary fibrosis140,141. In seem to affect the level of β‑catenin or its cate that the frequency in sporadic colorectal recent years, other mechanisms of activation cellular distribution. These drugs were cancers is actually closer to 1% (reviewed of the canonical Wnt pathway in tumours originally developed for the treatment of in Ref. 131). Over 3,500 different human have been discovered; for example, silencing other diseases and are approved by the US cancers were examined for the occurrence of the genes that encode the inhibitory Wnt Food and Drug Administration (FDA) of CTNNB1 mutations (colon cancer, ligands SFRPs and DKKs by hypermethyla‑ and European Medicines Agency, as pain melanoma, pilomatrixoma (hair tumours), tion, or by overexpression of Wnt proteins killers for example. NSAIDs can inhibit hepatocellular carcinoma, medulloblastoma, (WNT2B for example), Frizzleds (FZD10 colorectal tumours in rats and have cancer- hepatoblastoma, gastrointestinal tumours, for example) or Dishevelled142–149 (TIMELINE). preventive properities in epidemiological Wilms tumours132 and others), and in over Recently, activating LRP5 mutations were studies. Celecoxib has been approved for 700 cases mutations were found that were discovered in thyroid tumours150 (TIMELINE). the treatment of patients with FAP since predominantly centered in the N terminus Inappropriate mutation or deregulated 1999 (Ref. 163). In 2004, a high-throughput of CTNNB1 (Refs 121,133). The mutations expression of various genes of the Wnt path‑ screening method to search for low- in the N terminus either caused the dele‑ way are also the cause of many other diseases molecular-weight antagonists that target tion of an N‑terminal fragment (encoded that affect the cardiovascular, nervous, bone, the interaction between β‑catenin and by exon 3), or altered the N‑terminal kidney and other systems (BOX 1). TCF4 was reported, and fungal derivatives phosphorylation sites Ser45, Thr41, Ser37, It is now widely accepted that multiple were found that suppress the transcrip‑ Ser33 or neighbouring residues121 (FIG. 2b). If mutations are necessary for the development tional activity of β‑catenin164 (TIMELINE). exon 3 of Ctnnb1 is deleted by a conditional of human malignancy. In colon cancer, APC To our knowledge, these compounds have mutation in mice, adenomatous polyposis mutations represent early events in tumour not yet entered clinical trials, and it is not or other cancers develop134. Therefore, the progression (that is, they represent the clear whether their development has been tumour-causing mutations in APC, AXIN1, ‘gatekeeper’), but other mutations that affect carried further. The recent efforts of biotech AXIN2 and CTNNB1 generally lead to KRAS, SMAD2, SMAD4 and TP53 follow companies and the pharmaceutical industry inappropriate stabilization of β‑catenin. (see the Kinzler–Vogelstein model of the to develop effective inhibitors of the Wnt Surplus β‑catenin then translocates to the adenoma–carcinoma sequence in colorectal pathway for the treatment of patients with nucleus, interacts with the LEF and TCF cancer, reviewed in Refs 116,151–155). cancer and other diseases can be seen transcription factors, and persistently KRAS functions as a component of receptor on their webpages, such as The Genetics transactivates genes associated with the tyrosine kinases, and SMAD2 and SMAD4 Company, Nuvelo, Avalon and Curis.

Wnt SFRP Wnt Ligand Wnt Ligand DKK WIF

LRP5–6 Frizzled Notc RTKs ADAM h γ-secretas

adherin e

E-c PS1 P Gleevec CK1γ N C ? DVL β-catenin MAPK GSK3β NRTKs CK1α Axin GSK3β APC Low-molecular-weight MAPK interfering substances NSAIDs

Therapeutic/ recombinant proteins Antibodies

Pygo CBP ? BCL9 N C TARGETS Cele- COX2 coxib LEF/ TCF

Figure 3 | Cancer therapeutics that target components of the canon- a tyrosine kinase inhibitor and an approved therapeuticNatur fore chronicReviews myeloid| Cancer ical Wnt pathway. Schematic representation of canonical Wnt pathway leukaemia (CML), shows promise for the treatment of Wnt-induced gas- components that are already used as targets (red) or are future targets trointestinal tumours216,221. Interference with members of other signalling (green) for cancer therapeutics in animals and humans. The non-steroidal pathways, which directly or indirectly control β‑catenin stability (such as anti-inflammatory drug (NSAID) celecoxib, which targets the Wnt target γ‑secretase or ADAM, a disintegrin and metalloproteinase177,222) might also gene COX2 (cyclooxygenase 2), is already approved for the treatment of provide therapeutics against disease. APC, adenomatous polyposis coli; patients with familial adenomatous polyposis (FAP)163. New NSAIDs, thera- BCL9, B‑cell lymphoma 9; CBP, CREB-binding protein; CK1α, casein kinase peutic proteins, antibodies and low-molecular-weight products that inter- 1α; DKK, Dickkopf; DVL, Dishevelled; GSK3β, glycogen synthase kinase 3β; fere with aberrant activity of the canonical Wnt pathway are currently being LEF, lymphoid enhancer factor; LRP5, LDL-receptor related protein 5; MAPK, developed or might be future candidates. Transcriptional co-activators of mitogen-activated protein kinase; NRTKs, non-receptor tyrosine kinases; the Wnt pathway might also be future candidates for Wnt therapeutics. SFRP, secreted Frizzled-related protein; TCF, T‑cell factor; P, phosphoryla- Tissue-specific Wnt target genes might allow the development of drugs that tion; PS1, presenilin 1; Pygo, Pygopus; RTKs, receptor tyrosine kinases; WIF, target specific types of Wnt-induced tumour. Imatinib mesylate (Gleevec), Wnt inhibitory factor 1.

Recently, antibody-based therapies have also target transcriptional co-activators of the with BMP and Notch signalling in the been developed that target molecules of the Wnt pathway such as BCL9, CBP, CREB, intestinal stem cell niche to control stem cell Wnt pathway (such as Wnts or Frizzleds) BRG1, Pygopus, Hyrax and components of self-renewal179–181. Since then, Wnt signalling that are overexpressed in disease165,166. the Mediator complex are also under scrutiny has been shown to have an important role in Moreover, therapeutic proteins such as for potential therapeutic applications176,177 the stem cell compartments of various other SFRPs, which function as inhibitors of the (reviewed in Refs 161,162). However, it could tissues. We have shown that in the skin, loss- Wnt pathway, are presently being developed take a while until such novel low-molecular- of-function mutation of Ctnnb1 prevents the and tested in preclinical tumour models167 weight inhibitors and therapeutic proteins are generation of hair cell progenitors, but not (FIG. 3). TCF–β-catenin-restrictive oncolytic in the clinic. epidermal progenitors in the stem cell niche viruses are also under development168,169 of the follicular bulge182. This is in accordance (reviewed in Ref. 161). Wnt signalling and stem cells with the finding that activating mutations X‑ray structure analysis of many compo‑ Recent findings have demonstrated that the of Ctnnb1 in the skin lead to an expansion of nents of the Wnt pathway, such as β‑catenin, Wnt signalling pathway has an important hair precursor cells and are associated with axins, APC, TCFs, Dishevelled, BCL9 and role in the specification and maintenance of the formation of ‘hair tumours’: piloma‑ their complexes should allow the future precursor cell and stem cell lineages in various trixomas and trichofolliculomas183,184. In design and testing of low-molecular-weight tissues and organs. In 1998, the Clevers labo‑ addition, Wnt–β-catenin signalling has been substances that interfere with their activ‑ ratory reported the amazing finding that the proposed to increase haematopoietic stem ity, and many high-throughput screening mutation of the β‑catenin interaction partner cell renewal58,185–187. Here, Wnt signals also programmes to discover such compounds Tcf4 in mice resulted in the complete absence join forces with Notch signalling188. are running or planned164,170–175 (FIG. 3; of the stem cell compartment in the small Wnt–β-catenin signalling also regulates Supplementary information S1 (table)). intestine178 (TIMELINE). Further work has shown precursor cell maintenance in the central Low-molecular-weight compounds that that canonical Wnt signalling cooperates and peripheral nervous system189,190. For

example, the balance of neuronal progenitor a complete regression of the tumours202. Wnt pathway could make a great contribu‑ cell proliferation and patterning is disturbed These data demonstrate that Wnt–β-catenin tion to cures for cancer or other diseases. in the dorsal spinal cord of loss- and gain- signalling has an essential function in the There will soon be a broad spectrum of of-function Ctnnb1 mutant mice. We have maintenance of the mammary gland and tools to do this: therapeutic proteins (such shown that Wnt–β-catenin signalling in the skin cancer stem cells, and that the dif‑ as SFRPs and DKKs), nucleic-acid-based dorsal spinal cord is dependent on BMP ferentiation potential of β‑catenin in cancer substances (vaccines, interfering viruses and controls the action of the essential basic versus normal stem cells could be exploited and small interfering RNAs for example) helix-loop-helix transcription factor OLIG3 for therapy. and many low-molecular-weight interfering in the domain 2 and 3 postmitotic neu‑ It has also been shown that cytotoxic substances of Wnt pathway components. rons190. Recently, Wnt–β-catenin signalling drugs or irradiation often kill tumour cells, Effective therapies might necessitate inter‑ has also been shown to control self-renewal whereas putative cancer stem cells are resist‑ fering with protein–protein interactions and differentiation of Islet1-expressing ant (reviewed in Refs 203,204). It therefore at crucial steps of the pathway, which is precursor cells in neo- and postnatal hearts, needs to be determined whether the Wnt- still a great challenge for present-day drug which might be useful in the development inhibiting drugs that are presently being development215. New protein kinases that of cell-based therapies for regenerative developed (see above) have a preference for activate the Wnt pathway have recently medicine191,192 (reviewed in Ref. 193). Taken targeting cancer stem cells in the different been discovered96, and these hold great together, these data demonstrate that the tumour types. It has also been shown that promise for the development of specific Wnt pathway controls specification and the Wnt–β-catenin pathway is activated fur‑ inhibitors of these enzymes. Last but not maintenance of particular progenitor ther at late stages of tumour progression, for least, the combined use of inhibitors of and stem cell lineages in various tissues and example in colon carcinoma formation and different pathways will be examined, such organs during development and in the adult. metastasis155,205. Therefore, Wnt-inhibitory as inhibitors of receptor tyrosine kinases Recent results show that the Wnt pathway drugs might also interfere with metastasis and Notch, Hedgehog, or TGF‑β inhibitors is involved not only in development and formation and maintenance. in combination with Wnt inhibitors. For disease but also in regeneration and ageing example, tyrosine kinase inhibitors such as processes194–199. The next 25 years imatinib mesylate (Gleevec) and γ‑secretase The recently identified cancer stem cells In the past 25 years, many components and inhibitors perturb the Wnt signalling share many characteristics with normal target genes of the Wnt signalling pathway pathway (the Ras and the Notch pathways stem cells. These are the capacity for self- have been identified. Clearly the future will cooperate with the Wnt pathway in tumour renewal and differentiation into specific lead to the identification of new partners formation)216,217 (FIG. 3), and these will be cell types, as well as their dependence on a — we will gain a better understanding developed further. It might also be possible particular environment, the (cancer) stem of how components of the Wnt pathway to interfere at the target gene level. cell niche (reviewed in Ref. 200). A crucial function, and how they cooperate with the Recent studies suggest that so-called role of Wnt–β-catenin signalling in cancer many other proteins in cells206–209. We will cancer stem cells, which represent only a stem cells of the mammary gland and also gain a better understanding of how the minor fraction of tumour cells, might be epidermis has been identified. It was shown major signalling pathways interact during solely responsible for the generation and that the frequency of CD29+CD24+ stem development and why mutations in their maintenance of tumours. These cancer stem cells in the mammary gland increases by a components lead to tumour progression cells seem to share many characteristics with factor of 6.4 following activation of WNT1 and other diseases. A few specific unknown normal stem cells, including the capacity for (Ref. 201). Malanchi et al. have identified details that are likely to be resolved in the self-renewal and differentiation (see above). a population of stem cells in early mouse near future are, for example, the mechanism Perhaps these have a similar role in tumours. epidermal tumours that are characterized of action of BCL9 and PYGO in vertebrate This emerging concept provides exciting by phenotypic and functional similarities development and disease, the role of possibilities for both understanding tumour to normal follicular bulge epidermal stem tyrosine phosphorylation of different Wnt progression and possible therapeutic inter‑ cells202. In normal mouse skin, CD34+ fol‑ components in these processes, the roles of ference. Future research will therefore dem‑ licular bulge stem cells account for approxi‑ activating serine/threonine kinases in the onstrate how important the Wnt–β-catenin mately 1.8% of the keratinocytes. However, Wnt pathway and the actual mechanisms signalling pathway is for the self-renewal of in cutaneous tumours derived by chemical by which the Wnt pathway affects gene cancer stem cells. (dimethylbenzanthracene (DMBA) and regulation and chromatin remodelling The past two decades have seen a number 12-O-tetradecanoylphorbol-13-acetate (reviewed in Ref. 210). of cases in which processes that regulate (TPA)) carcinogenesis or by overexpression Over 90% of colon cancers and a high embryonic development have been impli‑ of mutant Ras (HRAS‑R12T59), a 9‑fold percentage of other cancers (reviewed in cated in disease. The ancient Wnt signalling increase of this CD34+ cell population Ref. 211) originate from activating muta‑ pathway controls many of the processes was seen. The tumorogenic capacity of tions in the Wnt pathway. Cell culture crucial for the growth, differentiation, the CD34+ cells was over 100-fold greater studies reveal that inhibition of the Wnt and regulation of animal cells, so it is little than that of the unsorted cells, and the pathway (by WIF1 or β‑catenin silenc‑ wonder that mutations in its components tumours resembled the architecture ing, RNA interference targeting GSK3β, have been linked to the deregulation of those of the parental tumours, including the Pygopus 2 antisense oligonucleotides or processes in disease. Learning to manipulate maintenance of a small population of the Vitamin D3, for example) can normalize this pathway through new low-molecular- CD34+ cells. Remarkably, Ctnnb1 deletion cancer cells — that is, inhibit prolifera‑ weight substances or recombinant or in DMBA–TPA or Ras-induced tumours tion and induce differentiation159,212–214. therapeutic molecules is of great promise for by conditional mutagenesis resulted in Therefore, it is likely that inhibiting the therapy in the future.

Alexandra Klaus and Walter Birchmeier are at the 23. Peifer, M., Sweeton, D., Casey, M. & Wieschaus, E. 49. Parker, D. S., Jemison, J. & Cadigan, K. M. Pygopus, Max Delbrück Centre for Molecular Medicine, Wingless signal and Zeste-white 3 kinase trigger a nuclear PHD-finger protein required for Wingless opposing changes in the intracellular distribution of signaling in Drosophila. Development 129, Robert-Roessle-Strasse 10, 13,125 Berlin, Germany. Armadillo. Development 120, 369–380 (1994). 2565–2576 (2002). Correspondence to W.B. 24. Siegfried, E., Wilder, E. L. & Perrimon, N. 50. Thompson, B., Townsley, F., Rosin-Arbesfeld, R., e-mail: [email protected] Components of wingless signalling in Drosophila. Musisi, H. & Bienz, M. A new nuclear component of the Nature 367, 76–80 (1994). Wnt signalling pathway. Nature Cell Biol. 4, 367–373 doi:10.1038/nrc2389 25. Peifer, M., Pai, L. M. & Casey, M. Phosphorylation of (2002). the Drosophila adherens junction protein Armadillo: 51. Brembeck, F. H. et al. Essential role of BCL9–2 in the 1. Nusse, R. & Varmus, H. E. Many tumors induced by roles for wingless signal and zeste-white 3 kinase. Dev. switch between β-catenin’s adhesive and transcriptional the mouse mammary tumor virus contain a provirus Biol. 166, 543–556 (1994). functions. Genes Dev. 18, 2225–2230 (2004). integrated in the same region of the host genome. 26. McMahon, A. P. & Moon, R. T. Ectopic expression of 52. Coudreuse, D. Y., Roel, G., Betist, M. C., Destree, O. & Cell 31, 99–109 (1982). the proto-oncogene int‑1 in Xenopus embryos leads to Korswagen, H. C. Wnt gradient formation requires 2. Lane, P. W. Swaying. Mouse News Lett. 36, 40 duplication of the embryonic axis. Cell 58, 1075–1084 retromer function in Wnt-producing cells. Science (1967). (1989). 312, 921–924 (2006). 3. Thomas, K. R. & Capecchi, M. R. Targeted disruption 27. Dominguez, I., Itoh, K. & Sokol, S. Y. Role of glycogen 53. Prasad, B. C. & Clark, S. G. Wnt signaling of the murine int‑1 proto-oncogene resulting in severe synthase kinase 3 β as a negative regulator of establishes anteroposterior neuronal polarity and abnormalities in midbrain and cerebellar dorsoventral axis formation in Xenopus embryos. requires retromer in C. elegans. Development 133, development. Nature 346, 847–850 (1990). Proc. Natl Acad. Sci. USA 92, 8498–8502 (1995). 1757–1766 (2006). 4. Thomas, K. R., Musci, T. S., Neumann, P. E. & 28. Guger, K. A. & Gumbiner, B. M. β-Catenin has Wnt-like 54. Kadowaki, T., Wilder, E., Klingensmith, J., Zachary, K. Capecchi, M. R. Swaying is a mutant allele of the activity and mimics the Nieuwkoop signaling center in & Perrimon, N. The segment polarity gene porcupine proto-oncogene Wnt‑1. Cell 67, 969–976 (1991). Xenopus dorsal–ventral patterning. Dev. Biol. 172, encodes a putative multitransmembrane protein 5. Sharma, R. P. Wingless, a new mutant in D. 115–125 (1995). involved in Wingless processing. Genes Dev. 10, melanogaster. Drosoph. Inf. Serv. 50, 134 (1973). 29. He, X., Saint-Jeannet, J. P., Woodgett, J. R., Varmus, 3116–3128 (1996). 6. Rijsewijk, F. et al. The Drosophila homolog of the H. E. & Dawid, I. B. Glycogen synthase kinase‑3 and 55. Haerry, T. E., Heslip, T. R., Marsh, J. L. & mouse mammary oncogene int‑1 is identical to the dorsoventral patterning in Xenopus embryos. Nature O’Connor, M. B. Defects in glucuronate biosynthesis segment polarity gene wingless. Cell 50, 649–657 374, 617–622 (1995). disrupt Wingless signaling in Drosophila. Development (1987). 30. Behrens, J. et al. Functional interaction of β-catenin 124, 3055–3064 (1997). 7. Cabrera, C. V., Alonso, M. C., Johnston, P., with the transcription factor LEF‑1. Nature 382, 56. Tanaka, K., Kitagawa, Y. & Kadowaki, T. Drosophila Phillips, R. G. & Lawrence, P. A. Phenocopies induced 638–642 (1996). segment polarity gene product porcupine stimulates with antisense RNA identify the wingless gene. Cell 31. Huber, O. et al. Nuclear localization of β-catenin by the posttranslational N‑glycosylation of wingless in 50, 659–663 (1987). interaction with transcription factor LEF‑1. Mech. Dev. the endoplasmic reticulum. J. Biol. Chem. 277, 8. Nüsslein-Volhard, C. & Wieschaus, E. Mutations 59, 3–10 (1996). 12816–12823 (2002). affecting segment number and polarity in Drosophila. 32. Molenaar, M. et al. XTcf‑3 transcription factor 57. Reichsman, F., Smith, L. & Cumberledge, S. Nature 287, 795–801 (1980). mediates β‑catenin‑induced axis formation in Xenopus Glycosaminoglycans can modulate extracellular 9. Nüsslein-Volhard, C., Wieschaus, E. & Kluding, H. embryos. Cell 86, 391–399 (1996). localization of the wingless protein and promote signal Mutations affecting the pattern of the larval cuticle in 33. Travis, A., Amsterdam, A., Belanger, C. & Grosschedl, R. transduction. J. Cell Biol. 135, 819–827 (1996). Drosophila melanogaster. I. Zygotic loci on the second LEF‑1, a gene encoding a lymphoid-specific protein with 58. Willert, K. et al. Wnt proteins are lipid-modified and can chromosome. Rouxs Arch. Dev. Biol. 193, 267–282 an HMG domain, regulates T‑cell receptor α enhancer act as stem cell growth factors. Nature 423, 448–452 (1984). function [corrected]. Genes Dev. 5, 880–894 (1991). (2003). 10. Nusse, R. et al. A new nomenclature for int‑1 and 34. van de, W. M., Oosterwegel, M., Dooijes, D. & 59. Kitajewski, J., Mason, J. O. & Varmus, H. E. related genes: the Wnt gene family. Cell 64, 231 Clevers, H. Identification and cloning of TCF‑1, a T Interaction of Wnt‑1 proteins with the binding protein (1991). lymphocyte-specific transcription factor containing a BiP. Mol. Cell. Biol. 12, 784–790 (1992). 11. Bittner, J. J. Some possible effects of nursing on the sequence-specific HMG box. EMBO J. 10, 123–132 60. Culi, J. & Mann, R. S. Boca, an endoplasmic reticulum mammary gland tumor incidence in mice. Science 84, (1991). protein required for wingless signaling and trafficking 162 (1936). 35. Waterman, M. L., Fischer, W. H. & Jones, K. A. A of LDL receptor family members in Drosophila. Cell 12. Spemann, H. & Mangold, H. Über Induktion von thymus-specific member of the HMG protein family 112, 343–354 (2003). Embryonalanlagen durch Implantation artfremder regulates the human T cell receptor C α enhancer. 61. Hsieh, J. C. et al. Mesd encodes an LRP5/6 chaperone Organisatoren. Wilhelm Roux Arch. Entw. Mech. Org. Genes Dev. 5, 656–669 (1991). essential for specification of mouse embryonic 100, 599–638 (1924) (in German). 36. Zeng, L. et al. The mouse Fused locus encodes Axin, an polarity. Cell 112, 355–367 (2003). 13. Glinka, A., Delius, H., Blumenstock, C. & Niehrs, C. inhibitor of the Wnt signaling pathway that regulates 62. Banziger, C. et al. Wntless, a conserved membrane Combinatorial signalling by Xwnt‑11 and Xnr3 in the embryonic axis formation. Cell 90, 181–192 (1997). protein dedicated to the secretion of Wnt proteins organizer epithelium. Mech. Dev. 60, 221–231 37. Behrens, J. et al. Functional interaction of an axin from signaling cells. Cell 125, 509–522 (2006). (1996). homolog, conductin, with β-catenin, APC, and GSK3β. 63. Bartscherer, K., Pelte, N., Ingelfinger, D. & 14. Morgan, T. H. The relation between normal and Science 280, 596–599 (1998). Boutros, M. Secretion of Wnt ligands requires Evi, abnormal development of the embryo of the frog, as 38. Itoh, K., Krupnik, V. E. & Sokol, S. Y. Axis a conserved transmembrane protein. Cell 125, determined by the effect of lithium chloride in solution. determination in Xenopus involves biochemical 523–533 (2006). Archiv fur Entwicklungsmechanik 16, 691–716 interactions of axin, glycogen synthase kinase 3 and β- 64. Goodman, R. M. et al. Sprinter: a novel (1902). catenin. Curr. Biol. 8, 591–594 (1998). transmembrane protein required for Wg secretion and 15. Klein, P. S. & Melton, D. A. A molecular mechanism 39. Huelsken, J. et al. Requirement for β-catenin in signaling. Development 133, 4901–4911 (2006). for the effect of lithium on development. Proc. Natl anterior-posterior axis formation in mice. J. Cell Biol. 65. Binari, R. C. et al. Genetic evidence that heparin-like Acad. Sci. USA 93, 8455–8459 (1996). 148, 567–578 (2000). glycosaminoglycans are involved in wingless signaling. 16. Perrimon, N. & Mahowald, A. P. Multiple functions of 40. Gluecksohn-Schoenheimer, S. The effects of a lethal Development 124, 2623–2632 (1997). segment polarity genes in Drosophila. Dev. Biol. 119, mutation responsible for duplications and twinning 66. Hacker, U., Lin, X. & Perrimon, N. The Drosophila 587–600 (1987). in mouse embryos. J. Exp. Zool. 110, 47–76 (1949). sugarless gene modulates Wingless signaling and 17. Perrimon, N., Engstrom, L. & Mahowald, A. P. Zygotic 41. Bhanot, P. et al. A new member of the frizzled family encodes an enzyme involved in polysaccharide lethals with specific maternal effect phenotypes in from Drosophila functions as a Wingless receptor. biosynthesis. Development 124, 3565–3573 (1997). Drosophila melanogaster. I. Loci on the X chromosome. Nature 382, 225–230 (1996). 67. Mikels, A. J. & Nusse, R. Wnts as ligands: processing, Genetics 121, 333–352 (1989). 42. Wang, Y. et al. A large family of putative transmembrane secretion and reception. Oncogene 25, 7461–7468 18. Wieschaus, E. & Riggleman, R. Autonomous receptors homologous to the product of the Drosophila (2006). requirements for the segment polarity gene armadillo tissue polarity gene frizzled. J. Biol. Chem. 271, 68. Hausmann, G., Banziger, C. & Basler, K. Helping during Drosophila embryogenesis. Cell 49, 177–184 4468–4476 (1996). Wingless take flight: how WNT proteins are secreted. (1987). 43. He, X. et al. A member of the Frizzled protein family Nature Rev. Mol. Cell Biol. 8, 331–336 (2007). 19. Riggleman, B., Schedl, P. & Wieschaus, E. Spatial mediating axis induction by Wnt-5A. Science 275, 69. Eaton, S. Retromer retrieves wntless. Dev. Cell 14, expression of the Drosophila segment polarity gene 1652–1654 (1997). 4–6 (2008). armadillo is posttranscriptionally regulated by 44. Pinson, K. I., Brennan, J., Monkley, S., Avery, B. J. & 70. Takada, R. et al. Monounsaturated fatty acid wingless. Cell 63, 549–560 (1990). Skarnes, W. C. An LDL‑receptor‑related protein mediates modification of Wnt protein: its role in Wnt secretion. 20. Klingensmith, J., Nusse, R. & Perrimon, N. The Wnt signalling in mice. Nature 407, 535–538 (2000). Dev. Cell 11, 791–801 (2006). Drosophila segment polarity gene dishevelled 45. Tamai, K. et al. LDL‑receptor‑related proteins in 71. Finch, P. W. et al. Purification and molecular cloning of encodes a novel protein required for response to Wnt signal transduction. Nature 407, 530–535 a secreted, Frizzled-related antagonist of Wnt action. the wingless signal. Genes Dev. 8, 118–130 (2000). Proc. Natl Acad. Sci. USA 94, 6770–6775 (1997). (1994). 46. Wehrli, M. et al. arrow encodes an LDL‑receptor‑related 72. Leyns, L., Bouwmeester, T., Kim, S. H., Piccolo, S. & 21. Noordermeer, J., Klingensmith, J., Perrimon, N. & protein essential for Wingless signalling. Nature 407, De Robertis, E. M. Frzb‑1 is a secreted antagonist of Nusse, R. Dishevelled and armadillo act in the 527–530 (2000). Wnt signaling expressed in the Spemann organizer. wingless signalling pathway in Drosophila. Nature 47. Belenkaya, T. Y. et al. pygopus Encodes a nuclear Cell 88, 747–756 (1997). 367, 80–83 (1994). protein essential for wingless/Wnt signaling. 73. Rattner, A. et al. A family of secreted proteins contains 22. Siegfried, E., Chou, T. B. & Perrimon, N. Wingless Development 129, 4089–4101 (2002). homology to the cysteine-rich ligand-binding domain signaling acts through zeste-white 3, the Drosophila 48. Kramps, T. et al. Wnt/wingless signaling requires of frizzled receptors. Proc. Natl Acad. Sci. USA 94, homolog of glycogen synthase kinase‑3, to regulate BCL9/legless-mediated recruitment of pygopus to the 2859–2863 (1997). engrailed and establish cell fate. Cell 71, 1167–1179 nuclear β‑catenin‑TCF complex. Cell 109, 47–60 74. Wang, S., Krinks, M., Lin, K., Luyten, F. P. & Moos, M., (1992). (2002). Jr. Frzb, a secreted protein expressed in the Spemann

organizer, binds and inhibits Wnt‑8. Cell 88, 757–766 101. Hecht, A., Vleminckx, K., Stemmler, M. P., van Roy, F. 127. Su, L. K. et al. Multiple intestinal neoplasia caused by (1997). & Kemler, R. The p300/CBP acetyltransferases a mutation in the murine homolog of the APC gene. 75. Glinka, A. et al. Dickkopf‑1 is a member of a new function as transcriptional coactivators of β-catenin in Science 256, 668–670 (1992). family of secreted proteins and functions in head vertebrates. EMBO J. 19, 1839–1850 (2000). 128. Korinek, V. et al. Constitutive transcriptional activation induction. Nature 391, 357–362 (1998). 102. Mosimann, C., Hausmann, G. & Basler, K. by a β‑catenin‑Tcf complex in APC–/– colon carcinoma. 76. Hsieh, J. C. et al. A new secreted protein that binds to Parafibromin/Hyrax activates Wnt/Wg target gene Science 275, 1784–1787 (1997). Wnt proteins and inhibits their activities. Nature 398, transcription by direct association with β-catenin/ 129. Morin, P. J. et al. Activation of β‑catenin–Tcf signaling 431–436 (1999). Armadillo. Cell 125, 327–341 (2006). in colon cancer by mutations in β-catenin or APC. 77. Mukhopadhyay, M. et al. Dickkopf1 is required for 103. Wolf, D., Rodova, M., Miska, E. A., Calvet, J. P. & Science 275, 1787–1790 (1997). embryonic head induction and limb morphogenesis in Kouzarides, T. Acetylation of β-catenin by CREB-binding 130. Rubinfeld, B. et al. Stabilization of β-catenin by the mouse. Dev. Cell 1, 423–434 (2001). protein (CBP). J. Biol. Chem. 277, 25562–25567 genetic defects in melanoma cell lines. Science 275, 78. Mao, B. et al. Kremen proteins are Dickkopf receptors (2002). 1790–1792 (1997). that regulate Wnt/β-catenin signalling. Nature 417, 104. de la Roche, M. & Bienz, M. Wingless-independent 131. Polakis, P. The many ways of Wnt in cancer. Curr. Opin. 664–667 (2002). association of Pygopus with dTCF target genes. Curr. Genet. Dev. 17, 45–51 (2007). 79. Kazanskaya, O. et al. R‑Spondin2 is a secreted Biol. 17, 556–561 (2007). 132. Koesters, R. et al. Mutational activation of the β- activator of Wnt/β-catenin signaling and is required for 105. Daniels, D. L. & Weis, W. I. β-catenin directly displaces catenin proto-oncogene is a common event in the Xenopus myogenesis. Dev. Cell 7, 525–534 (2004). Groucho/TLE repressors from Tcf/Lef in Wnt-mediated development of Wilms’ tumors. Cancer Res. 59, 80. Xu, Q. et al. Vascular development in the retina and transcription activation. Nature Struct. Mol. Biol. 12, 3880–3882 (1999). inner ear: control by Norrin and Frizzled‑4, a high- 364–371 (2005). 133. Schwarz-Romond, T. Diversin, eine neue Komponente affinity ligand-receptor pair. Cell 116, 883–895 106. He, T. C. et al. Identification of c‑MYC as a target of the des Wnt-Signalweges. Thesis, Free Univ. Berlin 1–89 (2004). APC pathway. Science 281, 1509–1512 (1998). (2002) (in German). 81. Munemitsu, S., Albert, I., Souza, B., Rubinfeld, B. & 107. Riese, J. et al. LEF‑1, a nuclear factor coordinating 134. Harada, N. et al. Intestinal polyposis in mice with a Polakis, P. Regulation of intracellular beta-catenin signaling inputs from wingless and decapentaplegic. dominant stable mutation of the β-catenin gene. levels by the adenomatous polyposis coli (APC) tumor- Cell 88, 777–787 (1997). EMBO J. 18, 5931–5942 (1999). suppressor protein. Proc. Natl Acad. Sci. USA 92, 108. Krasnow, R. E., Wong, L. L. & Adler, P. N. Dishevelled 135. Inomata, M., Ochiai, A., Akimoto, S., Kitano, S. & 3046–3050 (1995). is a component of the frizzled signaling pathway in Hirohashi, S. Alteration of β-catenin expression in 82. Peifer, M., Sweeton, D., Casey, M. & Wieschaus, E. Drosophila. Development 121, 4095–4102 (1995). colonic epithelial cells of familial adenomatous Wingless signal and Zeste-white 3 kinase trigger 109. Seifert, J. R. & Mlodzik, M. Frizzled/PCP signalling: a polyposis patients. Cancer Res. 56, 2213–2217 opposing changes in the intracellular distribution of conserved mechanism regulating cell polarity and (1996). Armadillo. Development 120, 369–380 (1994). directed motility. Nature Rev. Genet. 8, 126–138 136. Shtutman, M. et al. The cyclin D1 gene is a target of 83. van Leeuwen, F., Samos, C. H. & Nusse, R. Biological (2007). the β-catenin/LEF‑1 pathway. Proc. Natl Acad. Sci. activity of soluble wingless protein in cultured 110. Veeman, M. T., Axelrod, J. D. & Moon, R. T. A second USA 96, 5522–5527 (1999). Drosophila imaginal disc cells. Nature 368, 342–344 canon. Functions and mechanisms of β‑catenin‑ 137. Tetsu, O. & McCormick, F. β-catenin regulates (1994). independent Wnt signaling. Dev. Cell 5, 367–377 expression of cyclin D1 in colon carcinoma cells. 84. Peters, J. M., McKay, R. M., McKay, J. P. & Graff, J. M. (2003). Nature 398, 422–426 (1999). Casein kinase I transduces Wnt signals. Nature 401, 111. Weeraratna, A. T. et al. Wnt5a signaling directly 138. Major, M. B. et al. Wilms tumor suppressor WTX 345–350 (1999). affects cell motility and invasion of metastatic negatively regulates WNT/β-catenin signaling. Science 85. Sakanaka, C., Leong, P., Xu, L., Harrison, S. D. & melanoma. Cancer Cell 1, 279–288 (2002). 316, 1043–1046 (2007). Williams, L. T. Casein kinase iε in the wnt pathway: 112. Ouko, L., Ziegler, T. R., Gu, L. H., Eisenberg, L. M. & 139. Rivera, M. N. et al. An X chromosome gene, WTX, is regulation of β-catenin function. Proc. Natl Acad. Sci. Yang, V. W. Wnt11 signaling promotes proliferation, commonly inactivated in Wilms tumor. Science 315, USA 96, 12548–12552 (1999). transformation, and migration of IEC6 intestinal 642–645 (2007). 86. Amit, S. et al. Axin-mediated CKI phosphorylation of epithelial cells. J. Biol. Chem. 279, 26707–26715 140. Cheon, S. S. et al. β-Catenin stabilization dysregulates β-catenin at Ser 45: a molecular switch for the Wnt (2004). mesenchymal cell proliferation, motility, and pathway. Genes Dev. 16, 1066–1076 (2002). 113. Rubinfeld, B. et al. Association of the APC gene invasiveness and causes aggressive fibromatosis and 87. Liu, C. et al. Control of β-catenin phosphorylation/ product with β-catenin. Science 262, 1731–1734 hyperplastic cutaneous wounds. Proc. Natl Acad. Sci. degradation by a dual-kinase mechanism. Cell 108, (1993). USA 99, 6973–6978 (2002). 837–847 (2002). 114. Su, L. K., Vogelstein, B. & Kinzler, K. W. Association of 141. Moon, R. T., Kohn, A. D., De Ferrari, G. V. & 88. Yanagawa, S. et al. Casein kinase I phosphorylates the the APC tumor suppressor protein with catenins. Kaykas, A. WNT and β-catenin signalling: diseases Armadillo protein and induces its degradation in Science 262, 1734–1737 (1993). and therapies. Nature Rev. Genet. 5, 691–701 Drosophila. EMBO J. 21, 1733–1742 (2002). 115. Hulsken, J., Birchmeier, W. & Behrens, J. E‑cadherin (2004). 89. Aberle, H., Bauer, A., Stappert, J., Kispert, A. & and APC compete for the interaction with β-catenin 142. Vider, B. Z. et al. Evidence for the involvement of the Kemler, R. β-catenin is a target for the ubiquitin- and the cytoskeleton. J. Cell Biol. 127, 2061–2069 Wnt 2 gene in human colorectal cancer. Oncogene 12, proteasome pathway. EMBO J. 16, 3797–3804 (1994). 153–158 (1996). (1997). 116. Kinzler, K. W. & Vogelstein, B. Lessons from hereditary 143. Katoh, M., Kirikoshi, H., Terasaki, H. & Shiokawa, K. 90. Salomon, D. et al. Regulation of β-catenin levels and colorectal cancer. Cell 87, 159–170 (1996). WNT2B2 mRNA, up-regulated in primary gastric localization by overexpression of plakoglobin and 117. Bodmer, W. F. et al. Localization of the gene for cancer, is a positive regulator of the WNT– inhibition of the ubiquitin-proteasome system. J. Cell familial adenomatous polyposis on chromosome 5. β‑catenin–TCF signaling pathway. Biochem. Biophys. Biol. 139, 1325–1335 (1997). Nature 328, 614–616 (1987). Res. Commun. 289, 1093–1098 (2001). 91. Jiang, J. & Struhl, G. Regulation of the Hedgehog and 118. Leppert, M. et al. The gene for familial polyposis coli 144. Kirikoshi, H., Sekihara, H. & Katoh, M. Expression of Wingless signalling pathways by the F‑box/WD40- maps to the long arm of chromosome 5. Science 238, WNT14 and WNT14B mRNAs in human cancer, up- repeat protein Slimb. Nature 391, 493–496 (1998). 1411–1413 (1987). regulation of WNT14 by IFNγ and up-regulation of 92. Marikawa, Y. & Elinson, R. P. β-TrCP is a negative 119. Ashton-Rickardt, P. G. et al. High frequency of APC WNT14B by β-estradiol. Int. J. Oncol. 19, 1221–1225 regulator of Wnt/β-catenin signaling pathway and loss in sporadic colorectal carcinoma due to breaks (2001). dorsal axis formation in Xenopus embryos. Mech. Dev. clustered in 5q21–22. Oncogene 4, 1169–1174 145. Terasaki, H., Saitoh, T., Shiokawa, K. & Katoh, M. 77, 75–80 (1998). (1989). Frizzled‑10, up-regulated in primary colorectal cancer, 93. Cavallo, R. A. et al. Drosophila Tcf and Groucho 120. Groden, J. et al. Identification and characterization of is a positive regulator of the. Int. J. Mol. Med. 9, interact to repress Wingless signalling activity. Nature the familial adenomatous polyposis coli gene. Cell 66, 107–112 (2002). 395, 604–608 (1998). 589–600 (1991). 146. Okino, K. et al. Up-regulation and overproduction of 94. Levanon, D. et al. Transcriptional repression by AML1 121. Polakis, P. Wnt signaling and cancer. Genes Dev. 14, DVL‑1, the human counterpart of the Drosophila and LEF‑1 is mediated by the TLE/Groucho corepressors. 1837–1851 (2000). dishevelled gene, in cervical squamous cell carcinoma. Proc. Natl Acad. Sci. USA 95, 11590–11595 (1998). 122. Hart, M. J., de los, S. R., Albert, I. N., Rubinfeld, B. & Oncol. Rep. 10, 1219–1223 (2003). 95. Roose, J. et al. The Xenopus Wnt effector XTcf‑3 Polakis, P. Downregulation of β-catenin by human 147. Caldwell, G. M. et al. The Wnt antagonist sFRP1 in interacts with Groucho-related transcriptional Axin and its association with the APC tumor colorectal tumorigenesis. Cancer Res. 64, 883–888 repressors. Nature 395, 608–612 (1998). suppressor, β-catenin and GSK3 β. Curr. Biol. 8, (2004). 96. Davidson, G. et al. Casein kinase 1 γ couples Wnt 573–581 (1998). 148. Suzuki, H. et al. Epigenetic inactivation of SFRP genes receptor activation to cytoplasmic signal transduction. 123. Smits, R. et al. Apc1638T: a mouse model delineating allows constitutive WNT signaling in colorectal cancer. Nature 438, 867–872 (2005). critical domains of the adenomatous polyposis coli Nature Genet. 36, 417–422 (2004). 97. Zeng, X. et al. A dual-kinase mechanism for Wnt co- protein involved in tumorigenesis and development. 149. Sato, H. et al. Frequent epigenetic inactivation of receptor phosphorylation and activation. Nature 438, Genes Dev. 13, 1309–1321 (1999). DICKKOPF family genes in human gastrointestinal 873–877 (2005). 124. Liu, W. et al. Mutations in AXIN2 cause colorectal tumors. Carcinogenesis 28, 2459–2466 (2007). 98. Bilic, J. et al. Wnt induces LRP6 signalosomes and cancer with defective mismatch repair by activating 150. Bjorklund, P., Akerstrom, G. & Westin, G. An LRP5 promotes dishevelled-dependent LRP6 β-catenin/TCF signalling. Nature Genet. 26, 146–147 receptor with internal deletion in hyperparathyroid phosphorylation. Science 316, 1619–1622 (2007). (2000). tumors with implications for deregulated WNT/β- 99. Schwarz-Romond, T. et al. The DIX domain of 125. Satoh, S. et al. AXIN1 mutations in hepatocellular catenin signaling. PLoS. Med. 4, e328 (2007). Dishevelled confers Wnt signaling by dynamic carcinomas, and growth suppression in cancer cells by 151. Bos, J. L. et al. Prevalence of ras gene mutations in polymerization. Nature Struct. Mol. Biol. 14, 484–492 virus-mediated transfer of AXIN1. Nature Genet. 24, human colorectal cancers. Nature 327, 293–297 (2007). 245–250 (2000). (1987). 100. Yamamoto, H., Komekado, H. & Kikuchi, A. Caveolin is 126. Moser, A. R., Pitot, H. C. & Dove, W. F. A dominant 152. Forrester, K., Almoguera, C., Han, K., Grizzle, W. E. & necessary for Wnt‑3a‑dependent internalization of mutation that predisposes to multiple intestinal Perucho, M. Detection of high incidence of K‑ras LRP6 and accumulation of β-catenin. Dev. Cell 11, neoplasia in the mouse. Science 247, 322–324 oncogenes during human colon tumorigenesis. Nature 213–223 (2006). (1990). 327, 298–303 (1987).

153. Baker, S. J. et al. Chromosome 17 deletions and p53 180. He, X. C. et al. BMP signaling inhibits intestinal stem 207. DasGupta, R., Kaykas, A., Moon, R. T. & Perrimon, N. gene mutations in colorectal carcinomas. Science 244, cell self-renewal through suppression of Wnt–β‑catenin Functional genomic analysis of the Wnt–wingless 217–221 (1989). signaling. Nature Genet. 36, 1117–1121 (2004). signaling pathway. Science 308, 826–833 (2005). 154. Fearon, E. R. & Vogelstein, B. A genetic model for 181. van Es, J. H. et al. Notch/γ-secretase inhibition turns 208. Srahna, M. et al. A signaling network for patterning of colorectal tumorigenesis. Cell 61, 759–767 (1990). proliferative cells in intestinal crypts and adenomas neuronal connectivity in the Drosophila brain. PLoS 155. Fodde, R., Smits, R. & Clevers, H. APC, signal into goblet cells. Nature 435, 959–963 (2005). Biol. 4, e348 (2006). transduction and genetic instability in colorectal 182. Huelsken, J., Vogel, R., Erdmann, B., Cotsarelis, G. & 209. Stelzl, U. et al. A human protein–protein interaction cancer. Nature Rev. Cancer 1, 55–67 (2001). Birchmeier, W. β-Catenin controls hair follicle network: a resource for annotating the proteome. Cell 156. Kitamura, T. et al. SMAD4-deficient intestinal tumors morphogenesis and stem cell differentiation in the 122, 957–968 (2005). recruit CCR1+ myeloid cells that promote invasion. skin. Cell 105, 533–545 (2001). 210. Willert, K. & Jones, K. A. Wnt signaling: is the party in Nature Genet. 39, 467–475 (2007). 183. Gat, U., DasGupta, R., Degenstein, L. & Fuchs, E. the nucleus? Genes Dev. 20, 1394–1404 (2006). 157. van den Brink, G. R. et al. Indian Hedgehog is an De novo hair follicle morphogenesis and hair tumors in 211. Giles, R. H., van Es, J. H. & Clevers, H. Caught up in a antagonist of Wnt signaling in colonic epithelial cell mice expressing a truncated b-catenin in skin. Cell 95, Wnt storm: Wnt signaling in cancer. Biochim. Biophys. differentiation. Nature Genet. 36, 277–282 (2004). 605–614 (1998). Acta 1653, 1–24 (2003). 158. Reddy, B. S. et al. Chemoprevention of colon cancer by 184. Chan, E. F., Gat, U., McNiff, J. M. & Fuchs, E. 212. Shakoori, A. et al. Deregulated GSK3β activity in specific cyclooxygenase‑2 inhibitor, celecoxib, A common human skin tumour is caused by activating colorectal cancer: its association with tumor cell administered during different stages of carcinogenesis. mutations in β-catenin. Nature Genet. 21, 410–413 survival and proliferation. Biochem. Biophys. Res. Cancer Res. 60, 293–297 (2000). (1999). Commun. 334, 1365–1373 (2005). 159. Palmer, H. G. et al. Vitamin D(3) promotes the 185. Reya, T. et al. A role for Wnt signalling in self-renewal 213. Popadiuk, C. M. et al. Antisense suppression of differentiation of colon carcinoma cells by the of haematopoietic stem cells. Nature 423, 409–414 pygopus2 results in growth arrest of epithelial ovarian induction of E‑cadherin and the inhibition of β-catenin (2003). cancer. Clin. Cancer Res. 12, 2216–2223 (2006). signaling. J. Cell Biol. 154, 369–387 (2001). 186. Kirstetter, P., Anderson, K., Porse, B. T., 214. Lin, Y. C. et al. Wnt inhibitory factor‑1 gene transfer 160. Boon, E. M. et al. Sulindac targets nuclear β-catenin Jacobsen, S. E. & Nerlov, C. Activation of the canonical inhibits melanoma cell growth. Hum. Gene Ther. 18, accumulation and Wnt signalling in adenomas of Wnt pathway leads to loss of hematopoietic stem cell 379–386 (2007). patients with familial adenomatous polyposis and in repopulation and multilineage differentiation block. 215. Trosset, J. Y. et al. Inhibition of protein–protein human colorectal cancer cell lines. Br. J. Cancer 90, Nature Immunol. 7, 1048–1056 (2006). interactions: the discovery of druglike β-catenin 224–229 (2004). 187. Scheller, M. et al. Hematopoietic stem cell and inhibitors by combining virtual and biophysical 161. Dihlmann, S. & von Knebel, D. M. Wnt/β‑catenin‑ multilineage defects generated by constitutive β- screening. Proteins 64, 60–67 (2006). pathway as a molecular target for future anti-cancer catenin activation. Nature Immunol. 7, 1037–1047 216. Zhou, L. et al. Tyrosine kinase inhibitor STI‑571/ therapeutics. Int. J. Cancer 113, 515–524 (2005). (2006). Gleevec down-regulates the β-catenin signaling 162. Barker, N. & Clevers, H. Mining the Wnt pathway for 188. Duncan, A. W. et al. Integration of Notch and Wnt activity. Cancer Lett. 193, 161–170 (2003). cancer therapeutics. Nature Rev. Drug Discov. 5, signaling in hematopoietic stem cell maintenance. 217. Takahashi-Yanaga, F. & Sasaguri, T. The Wnt/β-catenin 997–1014 (2006). Nature Immunol. 6, 314–322 (2005). signaling pathway as a target in drug discovery. J. 163. Smigel, K. Arthritis drug approved for polyp 189. Lee, H. Y. et al. Instructive role of Wnt/β-catenin in Pharmacol. Sci. 104, 293–302 (2007). prevention blazes trail for other prevention trials. sensory fate specification in neural crest stem cells. 218. Schedl, A. Renal abnormalities and their J. Natl Cancer Inst. 92, 297–299 (2000). Science 303, 1020–1023 (2004). developmental origin. Nature Rev. Genet. 8, 791–802 164. Lepourcelet, M. et al. Small-molecule antagonists of 190. Zechner, D. et al. Bmp and Wnt/β-catenin signals (2007). the oncogenic Tcf/β-catenin protein complex. Cancer control expression of the transcription factor Olig3 219. Gaspar, C. & Fodde, R. APC dosage effects in Cell 5, 91–102 (2004). and the specification of spinal cord neurons. Dev. Biol. tumorigenesis and stem cell differentiation. Int. J. Dev. 165. You, L. et al. An anti‑Wnt‑2 monoclonal antibody 303, 181–190 (2007). Biol. 48, 377–386 (2004). induces apoptosis in malignant melanoma cells and 191. Klaus, A., Saga, Y., Taketo, M. M., Tzahor, E. & 220. Bienz, M. The subcellular destinations of APC inhibits tumor growth. Cancer Res. 64, 5385–5389 Birchmeier, W. Distinct roles of Wnt/β-catenin and proteins. Nature Rev. Mol. Cell Biol. 3, 328–338 (2004). Bmp signaling during early cardiogenesis. Proc. Natl (2002). 166. Nagayama, S. et al. Therapeutic potential of Acad. Sci. USA 104, 18531–18536 (2007). 221. Gleevec approved for first-line treatment of CML. FDA antibodies against FZD 10, a cell-surface protein, for 192. Qyang, Y. et al. The Renewal and differentiation of Consum. 37, 5 (2003). synovial sarcomas. Oncogene 24, 6201–6212 Isl1+ cardiovascular progenitors are controlled by a 222. Maretzky, T. et al. ADAM10 mediates E‑cadherin (2005). Wnt/β-catenin pathway. Cell Stem Cell 1, 1–15 (2007). shedding and regulates epithelial cell–cell adhesion, 167. DeAlmeida, V. I. et al. The soluble wnt receptor 193. Tzahor, E. Wnt/β-catenin signaling and cardiogenesis: migration, and β-catenin translocation. Proc. Natl Frizzled8CRD-hFc inhibits the growth of timing does matter. Dev. Cell 13, 10–13 (2007). Acad. Sci. USA 102, 9182–9187 (2005). teratocarcinomas in vivo. Cancer Res. 67, 5371–5379 194. Kawakami, Y. et al. Wnt/β-catenin signaling regulates 223. Nishisho, I. et al. Mutations of chromosome 5q21 (2007). vertebrate limb regeneration. Genes Dev. 20, genes in FAP and colorectal cancer patients. Science 168. Chen, R. H. & McCormick, F. Selective targeting to the 3232–3237 (2006). 253, 665–669 (1991). hyperactive β-catenin/T-cell factor pathway in colon 195. Stoick-Cooper, C. L. et al. Distinct Wnt signaling 224. McCrea, P. D., Turck, C. W. & Gumbiner, B. A homolog cancer cells. Cancer Res. 61, 4445–4449 (2001). pathways have opposing roles in appendage of the armadillo protein in Drosophila (plakoglobin) 169. Malerba, M., Daeffler, L., Rommelaere, J. & Iggo, R. D. regeneration. Development 134, 479–489 (2007). associated with E‑cadherin. Science 254, 1359–1361 Replicating parvoviruses that target colon cancer cells. 196. Brack, A. S. et al. Increased Wnt signaling during (1991). J. Virol. 77, 6683–6691 (2003). aging alters muscle stem cell fate and increases 225. Schneider, S., Steinbeisser, H., Warga, R. M. & 170. Graham, T. A., Weaver, C., Mao, F., Kimelman, D. & fibrosis. Science 317, 807–810 (2007). Hausen, P. β-catenin translocation into nuclei Xu, W. Crystal structure of a β-catenin/Tcf complex. 197. Liu, H. et al. Augmented Wnt signaling in a demarcates the dorsalizing centers in frog and fish Cell 103, 885–896 (2000). mammalian model of accelerated aging. Science 317, embryos. Mech. Dev. 57, 191–198 (1996). 171. Huber, A. H. & Weis, W. I. The structure of the β- 803–806 (2007). 226. Yost, C. et al. The axis-inducing activity, stability, and catenin/E-cadherin complex and the molecular basis of 198. Gurley, K. A., Rink, J. C. & Sanchez, A. A. β-catenin subcellular distribution of β-catenin is regulated in diverse ligand recognition by β-catenin. Cell 105, defines head versus tail identity during planarian Xenopus embryos by glycogen synthase kinase 3. 391–402 (2001). regeneration and homeostasis. Science 319, 323–327 Genes Dev. 10, 1443–1454 (1996). 172. Xing, Y., Clements, W. K., Kimelman, D. & Xu, W. (2008). 227. Huber, A. H., Nelson, W. J. & Weis, W. I. Three- Crystal structure of a β-catenin/axin complex suggests 199. Petersen, C. P. & Reddien, P. W. Smed‑βcatenin‑1 is dimensional structure of the armadillo repeat region a mechanism for the β-catenin destruction complex. required for anteroposterior blastema polarity in of β-catenin. Cell 90, 871–882 (1997). Genes Dev. 17, 2753–2764 (2003). planarian regeneration. Science 319, 327–330 228. Sakanaka, C., Leong, P., Xu, L., Harrison, S. D. & 173. Liu, J. et al. A small-molecule agonist of the Wnt (2008). Williams, L. T. Casein kinase iepsilon in the wnt signaling pathway. Angew. Chem. Int. Ed. Engl. 44, 200. Sneddon, J. B. & Werb, Z. Location, location, location: pathway: regulation of β-catenin function. Proc. Natl 1987–1990 (2005). the cancer stem cell niche. Cell Stem Cell 1, 607–611 Acad. Sci. USA 96, 12548–12552 (1999). 174. Shan, J., Shi, D. L., Wang, J. & Zheng, J. Identification (2007). 229. Hsu, W., Zeng, L. & Costantini, F. Identification of a of a specific inhibitor of the dishevelled PDZ domain. 201. Shackleton, M. et al. Generation of a functional domain of Axin that binds to the serine/threonine Biochemistry 44, 15495–15503 (2005). mammary gland from a single stem cell. Nature 439, protein phosphatase 2A and a self-binding domain. 175. Sampietro, J. et al. Crystal structure of a β-catenin/ 84–88 (2006). J. Biol. Chem. 274, 3439–3445 (1999). BCL9/Tcf4 complex. Mol. Cell 24, 293–300 (2006). 202. Malanchi, I. et al. Cutaneous cancer stem cell 230. Seeling, J. M. et al. Regulation of β-catenin signaling 176. Stadeli, R. & Basler, K. Dissecting nuclear Wingless maintenance is dependent on β-catenin signaling. by the B56 subunit of protein phosphatase 2A. signalling: recruitment of the transcriptional co- Nature 452, 650–653 (2008). Science 283, 2089–2091 (1999). activator Pygopus by a chain of adaptor proteins. 203. Lobo, N. A., Shimono, Y., Qian, D. & Clarke, M. F. The 231. Lin, X. & Perrimon, N. Dally cooperates with Mech. Dev. 122, 1171–1182 (2005). biology of cancer stem cells. Annu. Rev. Cell Dev. Biol. Drosophila Frizzled 2 to transduce Wingless signalling. 177. Teo, J. L., Ma, H., Nguyen, C., Lam, C. & Kahn, M. 23, 675–699 (2007). Nature 400, 281–284 (1999). Specific inhibition of CBP/β-catenin interaction rescues 204. Wang, J. C. & Dick, J. E. Cancer stem cells: lessons 232. Tsuda, M. et al. The cell-surface proteoglycan Dally defects in neuronal differentiation caused by a from leukemia. Trends Cell Biol. 15, 494–501 (2005). regulates Wingless signalling in Drosophila. Nature presenilin‑1 mutation. Proc. Natl Acad. Sci. USA 102, 205. Stein, U. et al. The metastasis-associated gene 400, 276–280 (1999). 12171–12176 (2005). S100A4 is a novel target of β-catenin/T-cell factor 233. Mao, J. et al. Low-density lipoprotein receptor- 178. Korinek, V. et al. Depletion of epithelial stem-cell signaling in colon cancer. Gastroenterology 131, related protein‑5 binds to Axin and regulates the compartments in the small intestine of mice lacking 1486–1500 (2006). canonical Wnt signaling pathway. Mol. Cell 7, Tcf‑4. Nature Genet. 19, 379–383 (1998). 206. Biemar, F. et al. Comprehensive identification of 801–809 (2001). 179. Haramis, A. P. et al. De novo crypt formation and Drosophila dorsal–ventral patterning genes using a 234. To, K. F. et al. Alterations of frizzled (FzE3) and juvenile polyposis on BMP inhibition in mouse whole-genome tiling array. Proc. Natl Acad. Sci. USA secreted frizzled related protein (hsFRP) expression in intestine. Science 303, 1684–1686 (2004). 103, 12763–12768 (2006). gastric cancer. Life Sci. 70, 483–489 (2001).

235. Ugolini, F. et al. WNT pathway and mammary Acknowledgements to tackle this disease. Raising awareness carcinogenesis: loss of expression of candidate We thank C. Birchmeier and R. Hodge for helpful discussions tumor suppressor gene SFRP1 in most invasive and improving our text. J. Fritzmann advised us on drugs and of this looming epidemic in Africa is carcinomas except of the medullary type. Oncogene clinical aspects. The work of our laboratory is funded by the the first step. If the international cancer 20, 5810–5817 (2001). German Research Foundation (DFG), the German Cancer Aid 236. Uematsu, K. et al. Activation of the Wnt pathway in (Deutsche Krebshilfe), and the German Federal Minister of community takes concerted action now, non small cell lung cancer: evidence of dishevelled Research and Technology (BMBF). working in partnership with the African overexpression. Oncogene 22, 7218–7221 (2003). 237. Uematsu, K. et al. Wnt pathway activation in DATABASES Health Ministries, another tragedy can be mesothelioma: evidence of Dishevelled overexpression Entrez Gene: prevented. To establish cancer care pro‑ and transcriptional activity of β-catenin. Cancer Res. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=gene 63, 4547–4551 (2003). Apc | armadillo | β‑TrCP | caveolin | CBP | CCL9 | CK1a | CTNNB1 | grammes in African countries requires the 238. Wissmann, C. et al. WIF1, a component of the Wnt dishevelled | DKK1 | FAM123B | FGF4 | FGF18 | Frizzled | FZD10 | integration of clinical and public-health pathway, is down-regulated in prostate, breast, lung, Hyrax | Indian hedgehog | Int1 | Islet1 | KRAS | Kremen | lef1 | and bladder cancer. J. Pathol. 201, 204–212 Legless | LRP5 | LRP6 | MYC | Norrin | NRCAM | OLIG3 | porcupine | systems so as to be truly comprehensive, (2003). Pygopus | SMAD2 | SMAD4 | TP53 | Ultrabithorax | VEGF | WIF1 | and must bring together prevention, early 239. Yoshikawa, S., McKinnon, R. D., Kokel, M. & WNT1 | WNT2B | Wnt5a | Wntless | zeste white 3 Thomas, J. B. Wnt-mediated axon guidance via National Cancer Institute: http://www.cancer.gov/ detection and diagnosis, treatment, pal‑ the Drosophila Derailed receptor. Nature 422, colon cancer | hepatocellular carcinoma | medulloblastoma | liative care and the investment needed to 583–588 (2003). melanoma | thyroid tumours | Wilms tumours 240. Tao, Q. et al. Maternal wnt11 activates the canonical deliver these services. This will require wnt signaling pathway required for axis formation in FURTHER INFORMATION trained staff, equipment, relevant drugs Xenopus embryos. Cell 120, 857–871 (2005). W. Birchmeier’s homepage: http://www.mdc-berlin.de/en/ 241. Li, X. et al. Sclerostin binds to LRP5/6 and research/research_teams/signal_transduction_invasion_ and information systems, supported by antagonizes canonical Wnt signaling. J. Biol. Chem. and_metastasis_of_epithelial_cells/index.html broad and effective partnerships between 280, 19883–19887 (2005). Avalon: http://www.avalonrx.com/content.aspx?id=36 242. Semenov, M., Tamai, K. & He, X. SOST is a ligand for Curis: http://www.curis.com/pipeline.php local health-care delivery systems, research LRP5/LRP6 and a Wnt signaling inhibitor. J. Biol. Nuvelo: http://www.nuvelo.com institutions, international organizations, Chem. 280, 26770–26775 (2005). Roel Nusse’s webpage: 243. Takeda, H. et al. Human sebaceous tumors harbor http://www.stanford.edu/~rnusse/wntwindow.html non-governmental organizations (NGOs), inactivating mutations in LEF1. Nature Med. 12, The Genetics Company: http://www.the-genetics.com/ national governments in developed and 395–397 (2006). ?menu=therapeutics&sub=wntinhibitors&doc=main 244. Martello, G. et al. MicroRNA control of Nodal developing countries, and the pharmaceuti‑ signalling. Nature 449, 183–188 (2007). SUPPLEMENTARY INFORMATION cal industry. The relevant organizations and 245. Barker, N. et al. Identification of stem cells in small See online article: S1 (table) intestine and colon by marker gene Lgr5. Nature 449, All links are active in the online pdf individuals must be brought together to 1003–1007 (2007). develop achievable and sustainable national cancer plans that are evidence-based, priority-driven and resource-appropriate for African countries. S c i e n ce a n d soc i ety Cancer burden The challenge of cancer control In 2002, 7.6 million people worldwide died of cancer. This was 13% of the global mor‑ tality burden and, perhaps surprisingly, in Africa more than the number of deaths from HIV/AIDS, TB and malaria combined Rebecca J. Lingwood, Peter Boyle, Alan Milburn, Twalib Ngoma, (~5.6 million)3 (FIG. 1a). John Arbuthnott, Ruth McCaffrey, Stewart H. Kerr and David J. Kerr The World Health Organization (WHO) has estimated that the global cancer burden Abstract | While the world is focused on controlling the spread of diseases such will increase, according to current trends, as HIV and malaria in the developing world, another approaching epidemic has from 10 million new cases per year in 2000 been largely overlooked. The World Heath Organization predicts that there will to 16 million in 2020. Remarkably, 70% of be 16 million new cancer cases per year in 2020 and 70% of these will be in the these cases will be in the developing world, rising from 5.2 million annually to 8.8 developing world. Many of these cancers are preventable, or treatable when million by 2020, an increase of ~60%. Sub- detected early enough. Establishing effective, affordable and workable cancer Saharan Africa will account for >1 million control plans in African countries is one step in the right direction toward of these cases by 2020 (REF. 4). limiting this epidemic. Although the AIDS epidemic has seen the relatively indolent tumour Kaposi sar‑ coma leap to the top of the cancer league In the developing world, one-third of developing world, the challenges posed are tables for Uganda, Swaziland, Malawi and cancers are potentially preventable and substantial (BOX 1). Zimbabwe, FIG. 2 shows the other prevalent another third are treatable if detected early1. The world is focused on controlling tumour types, with cervical and breast However, in many developing countries, the spread of HIV, tuberculosis (TB) and carcinoma predominating in women and governments and institutions face a wide malaria, which are all acknowledged to be prostate and liver cancer in men5. range of serious health problems and cancer major killers in the developing world, and is often not a priority in limited-resource huge sums of money are currently available Cancer infrastructure in Africa settings. Currently, a cancer diagnosis in to help combat these diseases2. Cancer One of the levers used to promote invest‑ the developing world means a painful and is set to become the newest epidemic in ment in cancer control in developed distressing death in most cases. Although the developing world, with the potential countries was the international comparison there is increasing awareness of the magni‑ to claim a vast number of lives, but cur‑ of relative spend and infrastructure in neigh‑ tude of the growing cancer problem in the rently there is limited funding available bouring nations, the lobbyists using these